View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Neurogenic Inflammation in Skin and Airways

Peter Baluk

Cardiovascular Research Institute, University of California, San Francisco, California, U.S.A.

Neurogenic inflanunation. in its original definition. permeability as a result of the focal. transient. and the plasma leakage induced by stimulation of periph­ fully reversible formation of gaps. approximately 0.5 eral sensory nerves. occurs in the postcapillary to 1.5 p.m in diameter. located in the intercellular venules of the skin and airways. Plasm" leakage is junctions of endothelial cells. The gaps can be visu­ accompanied by increased blood flow. which results alized by silver nitrate staining of the endothelial cell from dilatation of arterioles. In skin. these phenom­ borders. by lectin staining. or by scanning and trans­ ena are manifested as wheal and flare. respectively. mission electron microscopy. Neurogenic inflanuna­ Both phenomena are mediated by neuropeptides re­ tion can be inhibited by preventing the stimulation of leased from capsaicin-sensitive unmyelinated sensory sensory nerves. by presynaptic inhibition of neu­ nerve fibers. Substance P is the primary mediator ropeptide release from sensory nerves. or by blocking responsible for pla.ma leakage. acting via tachykinin neuropeptide receptors. The formation of endothe­ NK-l receptors. whereas both calcitonm. gene-re­ lial gaps can also be inhibited by anti-infl."mmatory lated peptide and substance P induce vasodilatation. drugs that stabilize endothelial cells. such as f3-ad­ Sensory nerve transmitters also cause release of his­ renergic agonists and steroids. Key words: substance tamine from mast cells. which contributes substan­ P/sensory nerves/plasma leakage/capsaicin. Journal of Inves­ tially to plasma leakage in the skin. but less so in the tigative Dermatology Symposium Proceedings 2:76-81, airways. Substance P causes an increase in vascular 1997

he goal of this chapter is to review the discovery of 1913). Bruce correctly proposed that the response was based on a neurogenic inflammation and to outline the mecha­ peripheral axon reflex (Fig 1). Lewis (1927) noted that the flare nisms that mediate plasma leakage and vasodilata­ component of the triple response in human skin, Le., the vasodi­ T tion. Examples of the nerves and blood vessels latation spreading around the original site of stimulation was involved in these phenomena are presented from the dependent on an intact sensory nerve supply and was abolished by skin and airways. Readers interested in learning more about local anesthesia. Lewis proposed the concept of sensory nerves with neurogenic inflammation are referred to several excellent reviews a 'nocifensor' or defense function many years before the sensory (ChaW, 1988; Maggi and Meli, 1988; Geppetti and Holzer, 1996; transmitters involved were chemically identified. Studies on neu­ Lundberg, 1996; McDonald, 1994a, 1997). rogenic plasma leakage, however, were not continued, and until the 1960s it was generally believed that the influence of the nervous HISTORY OF NEUROGENIC INFLAMMATION system on inflammation was restricted to vasodilatation (Lembeck, The four cardinal signs of inflammation recognized since antiquity 1983; Szolcsanyi, 1984). For example, in the 1950s evidence was are redness, swelling, heat, and pain. The first three signs clearly presented that the vasodilatation resulting from antidromic stimu­ involve blood vessels, resulting from vasodilatation, hyperemia, lation of sensory nerves was caused by ATP liberated from the and increased vascular permeability, whereas the fourth sign in­ nerves (Holton, 1959). V volves sensory ner es. Inflammation can be induced by a wide Interest in neurogenic inflammation was rekindled by the pio­ variety of stimuli, including direct mechanical damage of tissues, neering work of the Hungarian scientist Nicholas Jancs6, who exposure to bacterial toxins or exogenous chemical irritants, and by made several important discoveries and technical innovations: (i) the release of endogenous inflammatory mediators. He coined the term "neurogenic inflammation" to describe the Increased blood flow, as evidenced by redness and heat in the inflammatory response induced in the skin by direct electrical skin resulting from antidromic stimulation of the cut ends of stimulation of sensory nerves or by stimulation of the nerves by sensory nerves, has been known for more than a century. The chemical irritants, such as capsaicin, mustard oil, formalin, xylene, discovery of neurogenic inflammation as we know it today, how­ or hypertonic solutions (Jancs6, 1960; Jancs6 et aI, 1967). (ii) He ever, is usually credited to Ninian Bruce, although he did not use noted that these inflammatory stimuli produced not only vasodila­ this exact term for this phenomenon. He demonstrated that topical tation, but also increased vascular permeability, and that the effects application of irritants to the conjunctiva and skin of cats andrabbits were not blocked by cholinergic or adrenergic antagonists. (iii) He produced not only redness and heat, but also swelling, and that introduced the Evans blue method for quantifying plasma leakage these effects depended on an intact sensory nerve supply (Bruce, and a histologic method for labeling leaky blood vessels with colloidal pigments. (iv) Jancs6 introduced the use of capsaicin, the

Reprint requests to: Dr. Peter Baluk, Cardiovascular Research Institute, irritant constituent of hot peppers, as a specific tool to stimulate University of California, San Francisco, CA 94143-0130. sensory nerves, when administered in low doses, and to desensitize Abbreviations: CGRP, calcitonin gene-related peptide. or destroy them, when administered in higher doses. After desen-

1087-0024/97/$10.50 • Copyright © 1997 by The Society fo r Investigative Dennatology, Inc.

76 VOL. 2. NO. 1 AUGUST 1997 NEUROGENIC INFLAMMATION IN SKIN AND AIRWAYS 77

facial Sk n nose tootn pu.Pi

_ -- _ - conJunctl"a 'Po1Tf.. ;ar 'R,oo't &0"131;°" erll

front paws

Figure 1. Diagrant illustrating the concept of an axon reflex­ ntediating antidrontic vasodilatation in the skin. [From Bruce (1913). esophagus Vaso-dilator axon reflexes. Q J Exp Physiol 6:339-354.]

u eter s b radder sitization with capsaicin, antidromic nerve stimulation or sensory pro s tate nerve irritants failed to induce inflammation or pain. (v) He noted that after stimulation with low doses of capsaicin, the pain threshold of sensory nerve fibers was decreased, i.e., the fibers became sensitized and the individuals developed hyperalgesia. (vi) Jancso clearly recognized that some inflammatory mediators, e.g., capsa­ icin, acted through a neurogenic mechanism, whereas others, e.g., histamine, acted through a non-neurogenic mechanism. Jancso proposed that for "many inflammation-causing substances, ... the inflammation is mediated entirely by the nervous system," and the joint s "pain sensory nerve endings are actually the receptors of {the} inflammatory reaction" Oancso, 1960). Thus, Jancso laid the basis Figure 2. Schentatic diagrant of the ntajor sites of neurogenic for future studies of the efferent effects of afferent nerves. (vii) He inflantntation in the rat, as evidenced by the leakage of plasnta proposed that a neurotransmitter or a "neurohumor" was released containing Evans blue dye. from the sensory nerve terminals, but its chemical identity eluded him. Unfortunately, the significance of Jancso's work was only generally recognized after his death in 1966. For fascinating accounts of the early studies of neurogenic inflammation by Jancso all likely to be caused by the same mechanisms. There has been

and others, see Jancso (1960), Jancso et al (1967), Szolcsanyi (1984), much interest in the possible role of sensory nerve pep tides in and Chahl (1988). bronchoconstriction, a characteristic of asthma (Barnes, 1996). It is becoming apparent that, in addition to the direct effects mentioned WHICH PHENOMENA CHARACTERIZE NEUROGENIC above, sensory neuropeptides have other longer term trophic and INFLAMMATION? mitogenic effects on endothelial cells, connective tissue cells, and For Jancso, the most characteristic sign of neurogenic inflammation immune ce�s mediated via a variety of receptors and intermediates

was clearly the increased vascular permeability that resulted in (payan, 1992; Pincelli et ai, 1993; Ziche, 1996; Ansel, 1997). In plasma leakage and edema. Although he noted that vasodilatation view of the diversity of these phenomena, it may be best to keep and hyperemia accompanied the increase in vascular permeability, our definitions clear and to restrict the term "neurogenic inflam­ he recognized that the former and the latter were separate phe­ mation" to the original concept of neurogenically induced plasma nomena, occurring in arterioles and in small venules, respectively. leakage as proposed by Jancso. Vasodilatation could also occur under many other circumstances WHERE DOES NEUROGENIC INFLAMMATION OCCUR? independently of plasma leakage. Jancso also made the important observation that the vasodilatation in skin was abolished by local Inflammatory changes mediated by sensory nerves have been anesthetics, but the neurogenic plasma leakage was not, suggesting reported in many tissues and organs (Fig 2). These include the skin, that these phenomena involved different mechanisms. The purpose conjunctiva, meninges and peripheral nerve sheaths, tympanic of neurogenic plasma leakage and vasodilatation is believed to be membrane, gingiva, dental pulp, tongue, salivary gland ducts, nasal protective. Increased local blood flow may be sufficient to sweep mucosa and airways, esophagus, stomach, gall bladder, bile and away mild irritants, whereas extravasated plasma may help to dilute pancreatic ducts, ureters and urinary bladder, female reproductive

or neutralize stronger irritants and flush them away into lymphatics tract, anal mucosa, and joints (Saria et ai, 1983; Geppetti and or the airway lumen. If the stimulus is not successfully eliminated, Holzer, 1996). The common features of many, but not all, of these however, the inflammatory process may become chronic and tissues is that they are surfaces of the body facing the external contribute to pathophysiology, as may be the case in asthma environment, which are densely innervated by chemosensitive (Barnes, 1996). sensory nerves. Another feature is that large differences in sensitiv­ Since Jancso's studies, many new effects of sensory nerve trans­ ity occur between regions and among individuals. Neurogenic mitters have been discovered, and popular definitions of neuro­ inflammation appears not to occur in the central nervous system, genic inflammation have expanded from the original concept of most of the gastrointestinal tract, heart, or in skeletal muscle (Saria

increased plasma leakage and have become somewhat blurred. et ai, 1983). Depending on the organ concerned, in addition to vasodilatation There are also significant differences in the magnitude and and sensory nerve hypersensitivity, other phenomena that have distribution of neurogenic inflammation among different species, become loosely associated with neurogenic inflammation include with the phenomenon particularly conspicuous and well studied in smooth muscle contraction, epithelial cell secretion and ion trans­ rats, guinea pigs, and mice (Fig 3). Neurogenic plasma leakage has port, mast cell degranulation, and the recruitment, vascular adhe­ also been reported in the skin of cats, rabbits, pigs, and humans. sion, activation, or migration of leukocytes (McDonald, 1994a). Although neurogenic plasma inflammation is very obvious in Clearly, not all of these phenomena occur in the skin, nor are they rodent airways, it is still not clear whether it occurs in human 78 BALUK JID SYMPOSIUM PROCEEDINGS

Figure 3. Leaky blood vessels inwhole mount preparadons of rat skin and trachea. (A-C) Skin of rat hindpaw after topicalapplication of an irritant (5% mustard oil). The particulate tracer Monastral blue is' trapped in the basallarnina of leaky blood vessels. (A) Low magnification view of labeled blood

vessels, mostly post-capillary venules (�) . Scale bar, 100 p.m. (B) Bundles of varicose substance P-immunoreactive nerve fibers (�) in proximity to Monastral blue-labeled blood vessels. Scale bar, 50 /Lm. (C) Mast cells (globular cells intensely stained with toluidine blue indicated by �) in proximity to leaky venules labeled with Monastral blue. Note the poor correspondence between the position of the mast cells and the intensity of leakage in nearby vessel segments. Scale bar, 50 /Lm. (D-F) Venules in rat trachea after intravenous injection of substance P to cause inflammation. (D) Immunohistochemical localization of

NK-1 receptors internalized into small dot-like endosomes (�) within the cytoplasm of inflamed venular endothelial cells. Scale bar, 10 /Lm. (B)Mter staining with silver nitrate, endothelial cell borders are visualized as solid black lines, and gaps are revealed as dot-like silver deposits (�) in leaky venules. Scale bar, 10 /Lm. (F) The border between parts of two endothelial cells (top and bottom of photograph) observed en fa ce from the luminal surface by scanning electron microscopy. Gaps (�) between the endothelial cells are separated by long endothelial cytoplasmic fingers. The underlying basallarnina is visible through some of the gaps. Scale bar, 1 /Lm. [From Baluk et al (1997).] VOL. 2, NO. 1 AUGUST 1997 NEUROGENIC INFLAMMATION IN SKIN AND AIRWAYS 79

Table I. Stimuli That Activate Sensory Nerves in produces plasma extravasation and vasodilatation that mimics the Neurogenic InB.lIIIlIDation electrical or chemical stimulation of the sensory nerves (Lembeck and Holzer, 1979); (iv) Depletion of substance P by pre-treatment 1. Diret:t activators: electrical stimulation, high K+, low pH, mechanical with high doses of capsaicin, or blocking the effect of substance P stimuli, heat, ischemia 2. Plant extracts: capsaicin, resiniferatoxin, mustard oil, nicotine with receptor antagonists, blocks the effect of sensory nerve 3. I'!f/ammatory mediators: histamine, bradykinin, serotonin, arachidonic stimulation. (v) Inhibiting the enzymes that naturally degrade acid metabolites substance P (see below) potentiates the effect of nerve stimulation. 4. Noxiaus vapors: cigarette smoke, formaldehyde, acrolein, mustard oil Intradermal injection of substance P into the human forearm causes vapor, xylene, ether, toluene diisocyanate, allergens, sulfur dioxide, a dose-dependent wheal and Bare response (Foreman et aI, 1983). cold dry air Subsequent to the identification of substance P as the mediator of neurogenic inflanunation, it was discovered that other structurally related tachykinins called neurokinin A and neurokinin B, and an unrelated peptide calcitonin gene-related peptide (CGRP), were airways (McDonald. 1997). Topical application of irritant chemicals co-localized with substance P within sensory neurons. All of these such as capsaicin, xylene, mustard oil, or even a piece of very sensory neuropeptides are co-released from sensory nerve endings pungent paprika to human skin produces a strong burning sensa­ by appropriate stimuli (Hua and Yaksh, 1992; Lou, 1993). All three tion, slight edema (wheal), and a flush at the site of application, tachykinins evoke plasma leakage if administered in high enough surrounded by an intense flare. Neurogenic inflammation does not concentrations, but substance P is more potent than neurokinin A, occur in denervated human skin, e.g., in patients who have and neurokinin B is not present in significant amounts in peripheral sustained sensory nerve injury Oancs6 et aI, 1967, 1968). sensory nerves (Chahl, 1988). Substance P, neurokinin A, and SENSORY NERVES THAT EVOKE NEUROGENIC neurokinin B selectively activate three distinct neurokinin receptors INFLAMMATION termed NK-1, NK-2, and NK-3 receptors. The use of selective neurokinin receptor agonists and antagonists has established that Anatomically, the sensory nerves that mediate antidromic vasodi­ the plasma leakage and vasodilatation induced by tachykinins are latation and neurogenic plasma extravasation are small diameter both mediated by NK-1 receptors (Lundberg, 1996; Maggi et aI, (A-8) myelinated, and unmyelinated (C) primary aiferent nerve 1993). NK-1 receptors have been localized and quantified in fibers. The (small dark B-type) cell bodies of these primary aiferent venular endothelial cells at sites of inflanunation (Bowden et aI, nerve fibers are located in the dorsal root ganglia and in the ganglia 1994 and Fig 3D). The biology of the NK-1 receptors is discussed of cranial nerves. The peripheral nerve fibers are distributed in more detail in the article in this volume by Grady et al (1997). somatotopically in the skin with sharply demarcated receptive fields CGRP by itself does not produce plasma leakage, but does (Pinter and Szolcsanyi, 1995). The skin and the airways are richly supplied by nerve fibers that contain neuropeptides and other induce vasodilatation, acting via CGRP 1 receptors (Brain and transmitters (Wallengren et aI, 1987; Lundberg et aI, 1988; Karanth Williams, 1985; Lundberg, 1996). CGRP, however, potentiates the et aI, 1991; Brain, 1996). The number and distribution of sensory plasma leakage induced by tachykinins in the skin and in the nerves and their receptors in the skin can change in various airways (Brain and Williams, 1985; Gamse and Saria, 1985), most pathologic conditions, such as psoriasis (Pincelli et aI, 1993). likely by increasing the blood flow. In this respect, CGRP is probably not di1ferent from other vasodilators, and the effect is not MEDIATORS OF NEUROGENIC PLASMA LEAKAGE AND restricted to neurogenic inflammation. For instance, prostaglandin VASODILATATION El induces hyperemia in guinea pig and rabbit skin and potentiates Substance P was suspected as a major neurotransmitter of primary the plasma leakage caused by bradykinin (Kopaniak et aI, 1978). a1£erent sensory neurons for many years after its discovery by von The plasma leakage induced in rat skin by substance P is potentiated Euler and Gaddum in the 1930s (see Chahl, 1988). This was by prostaglandin El and ATP (Chahl, 1977). In addition to its confirmed in the 1970s and 1980s, when substance P was isolated effects on blood vessels, CGRP has trophic effects on various cells and characterized. and antibodies to it became available. Substance (Ziche, 1996). P satisfies all of the criteria necessary for establishing a candidate A special feature of some unmyelinated aiferent nerve fibers is neurotransmitter as the mediator of neurogenic inflammation that they are polymodal, i.e., they can respond to a wide range of (Lembeck and Holzer, 1979): (i) It is present in the cell bodies and irritant stimuli listed in Table I. Other unmyelinated aiferent nerve nerve fibers of a population of primary aiferent neurons; (ii) It is fibers, however, respond selectively to low threshold thermal or released from sensory nerve endings upon electrical or chemical mechanical stimuli (Belmonte and Cervero, 1996). A further stimulation of the nerves. In many tissues, release of substance P is distinguishing feature of polymodal sensory nerves is that in quite difficult to demonstrate, because endogenous levels are so response to stimulation they transmit action potentials both ortho­ low, or it is rapidly degraded by enzymes. Tachykinin release, dromically to the central nervous system, and antidromically, however, has been shown clearly in the trachea and dental pulp, causing the release of transmitters at their peripheral ends (Maggi which contain high levels of substance P (Hua and Yaksh, 1992; and Meli, 1988; Lundberg, 1996; Maggi, 1996). The dual aiferent­ Ohkubo et ai, 1993); (iii) Administration of exogenous substance P efferent nature of the sensory nerves is responsible for the action of

Table D. Comparison of Plasma Leakage and Vasodilatation in Neurogenic Inflammation in Skin

Feature Plasma Leakage Vasodilatation References

Physical manifestation Wheat or 'tumor' Flare or 'rubor'

Area alfected Localized Widespread (Foreman et aI, 1983) Blood vessels involved Postcapillary venules Precapillary arterioles Oancs6 et aI, 1967) Nerve mechanism 'Axon response' 'Axon reflex' (Szolcsanyi, 1988, 1996) Sensory nerves involved Capsaicin-sensitive C-fibers Capsaicin-sensitive A-8 and C-fibers Oiinig and Lisney, 1989) Time course to maximum Long (12 min) Short (3 min) (Foreman et aI, 1983) Neurotransmitters Substance P Substance P (Lembeck and Holzer, 1979) CGRP (Brain and Williams, 1985)

Inhibition by H1 antihistamines Small (27%) Large (78%) (Foreman et aI, 1983) Mect of local anesthetics Resistant Inhibited (Maggi and Meli, 1988; Szolcsanyi, 1996) Mect of tetrodotoxin Resistant Inhibited (Maggi and Meli, 1988; Szolcsanyi, 1996) 80 BALUK JID SYMPOSIUM PROCEEDINGS

Table m. Drugs That Inhibit Neurogenic Plasma approximately one order of magnitude lower for vasodilatation Leakage and Vasodilatation" than for plasma leakage (Lembeck and Holzer, 1979). Antidromic electrical stimulation of sensory nerves can be prevented by local 1. Desensitization of sensory nerves or depletion of transmitters anesthetics, such as lidocaine, that block axonal conduction. Inter­ Chronic treatment with capsaicin or resiniferatoxin estingly, these same anesthetics do not block the release of trans­ 2. Prevention of sensory nerve activation Capsazepine (by blocking capsaicin receptors) mitter induced by the direct stimulation of sensory nerve endings Ruthenium red (by blocking ion channels) (Jancso et aI, 1967; Szolcsanyi, 1996). Consequently, local anes­ Local anesthetics (by preventing nerve conduction) thetics block vasodilatation, which requires an axon rtiflex involving

Sodium cromoglycate (by stabilizing mast cells) nerve conduction, but not plasma leakage, which requires only an 3. Presynaptic inhibition of transmitter release axon response involving the local release of transmitter from nerve Opioids acting via lJr-opioid receptors fibers but not nerve conduction (Szolcsanyi, 1996). The similarities Catecholamines acting via u2-adrenoceptors and differences between neurogenic plasma leakage and antidromic Neuropeptide Y acting via Y2 receptors vasodilatation are summarized in Table II. GABA acting via GABA2 receptors Histamine acting via H3 receptors ROLE OF MAST CELLS AND HISTAMINE IN Corticotropin releasing fa ctor and related peptides NEUROGENIC INFLAMMATION Potassium channel openers 4. Receptor antagonists Mast cells contribute to neurogenic plasma leakage and vasodila­ NK-1 receptor antagonists tation in some, but not all, tissues where these phenomena have CGRP1 receptor antagonists been reported and are abundant in skin and airways. In neurogenic Histamine H1 receptor antagonists inflammation, these actions of mast cells are triggered by release of 5. Anti-inflammatory drugs that stabilize endothelial cells transmitters from sensory nerves, mainly substance P (Foreman et Steroids aI, 1983). Although some mast cells are located close to leaky blood J3-Adrenergic agonists vessels or to sensory nerves, other mast cells are distant from either Phosphodiesterase inhibitors and other cAMP-inducing agents blood vessels or sensory nerves (Fig 3C). It is not clear from what • For further details see: Barnes (1996), Barnes et al (1991), Lundberg (1996) and distance peptides released from sensory nerves can evoke the Maggi and Meli (1988). release of mediators from mast cells and vice versa. Many other non-neurogenic stimuli, such as compound 48/80 or allergen exposure, can also cause mast cell degranulation. Apparently, these nerves on blood vessels (Maggi and Meli, 1988). Irritant substance P does not induce the release of histamine via NK-1, stimuli can act on nerves either directly or indirectly. For instance, NK-2, or NK-3 receptors. Instead, the charged N-terminus of the bradykinin acts mainly by inducing the production of prostaglandin substance P molecule has a direct effect on mast cell intracellular intermediates (Geppetti et aI, 1990). Some cytokines, e.g., interleu­ G-proteins (Mousli et aI, 1990). In human skin, the wheal and flare kin 1{3, can potentiate neurogenic inflammation by lowering the responses to intradermal injections of substance P are both reduced, threshold of sensory nerves to irritant stimuli, i.e., by inducing but are not entirely eliminated by pre-treatment with H1 histamine hyperalgesia (Herbert and Holzer, 1994; Lundberg, 1996). antagonists (Foreman et aI, 1983). Thus, it appears that substance P Capsaicin deserves special consideration among irritant stimuli, has both a direct action on blood vessels and an indirect one, via not only for historical reasons, but because it has been used as a tool histamine release, from mast cells. Conversely, histamine can also to dljine the sensory nerves that cause neurogenic inflammation evoke plasma leakage by acting directly on endothelial cells (Majno (Lundberg, 1996; Maggi, 1996). Capsaicin depolarizes the capsa­ and Palade, 1961). icin-sensitive nerves at low concentrations, probably by activating a Marked differences in the extent of mast cell involvement in specific capsaicin receptor and causing a nonselective influx of neurogenic inflammation exist among different organs and species cations. The binding of capsaicin to its receptor and the nonspecific and even among different strains of the same species (Chahl, 1988). cation channel can be blocked by capsazepine and ruthenium red, Two examples will suffice: (i) in response to antidromic electrical respectively (Lundberg, 1996; Maggi, 1996). At higher concentra­ nerve stimulation or injection of substance P, plasma leakage is tions, capsaicin first produces desensitization of the nerves and, increased in both the lower lip and in the tooth pulp of rats, but after chronic application, it causes permanent structural damage. In mast cells are involved only in the lower lip (Ohkubo et ai, 1993): adult animals, only the sensory nerve endings are depleted of (ii) mast cells play a role in neurogenic inflammation of the airways transmitter, but in newborn animals, entire sensory neurons can be of some rat strains, but not in others (Germonpre et ai, 1995). permanently destroyed (Jancso et aI, 1985). In both cases there is MODULATION OF NEUROGENIC PLASMA LEAKAGE desensitization to further irritant stimuli that act via capsaicin­ AND VASODILATATION sensitive nerves, but there is no effect on the inflammation caused by mediators that act directly on blood vessels, or on the perception Because of the obvious therapeutic potential, much attention has of mechanical stimuli via other nerves that are not sensitive to been devoted to regulating neurogenic inflammation. In principle, capsaicin (Jancso et aI, 1967). Resiniferatoxin, the ultrapotent the inflammatory effects of sensory neuropeptides can be potenti­ analog of capsaicin, has a similar, but not identical, spectrum of ated or inhibited in several ways as listed in Table III. Presynaptic effects and is reviewed elsewhere in this issue by Biro et al (1997). inhibition of sensory neuropeptide release can be induced by seemingly diverse molecules that hyperpolarize sensory nerve fibers COMPARISON OF NEUROGENIC PLASMA LEAKAGE by various ion channels (Barnes et ai, 1991). An interesting question AND VASODILATATION is whether endogenous anti-inflammatory substances exist that As noted by Jancso, although neurogenic plasma leakage and inhibit sensory peptide release and, ifso, whether they can be used vasodilatation often occur together and are triggered by the same therapeutically. A potential advantage of such treatment would be stimuli, they are clearly separate and independent phenomena the ability to reduce the inflammation induced by many different involving different parts of the microvasculature. Antidromic stim­ stimuli. Preventing the release of multiple sensory nerve pep tides ulation of sensory nerves at an intensity of 10 V or more and presynaptically may prove to be more effective than blocking the frequencies of 1-1 0 Hz induces release of pep tides from the sensory effects of individual peptides postsynaptically by using multiple nerve fibers (Jancso et aI, 1967; Lundberg and Saria, 1983; Mc­ receptor antagonists. Donald et aI, 1988) suggesting that A-a (thinly myelinated) and LEAKY BLOOD VESSELS AND ENDOTHELIAL GAPS C-fibers (unmyelinated) axons are involved. In the skin A-a fibers are preferentially responsible for vasodilatation, whereas the C­ The blood vessels that leak plasma in neurogenic inflammation are fibers are responsible for plasma leakage (Jaoig and Lisney, 1989). predominantly postcapillary venules and collecting venules; arte­ In rat skin, the threshold to intravenous injection of substance P is rioles and capillaries do not become leaky (Jancso et aI, 1967). The VOL. 2, NO. 1 AUGUST 1997 NEUROGENIC INFLAMMATION IN SKIN AND AIRWAYS 81

endothelial basement membrane of leaky venules can be labeled Jancso N: Role of thenerve terminalsin themechanism ofinft ammatoryreactions. BlIlI NY) 7:53--77, histologically by the entrapment of intravenously injected particu­ Millard Fillmore Hosp (Buffalo, 1960 J ancsoN, J ancs6-Gabor A, SzolcsanyiJ: Direct evidence for neurogenic inflammation late tracers such as colloidal silver or carbon or Monastral blue (Fig and its prevention by denervation and by pretreatment with capsaicin. Br J 3A-q. The pattern of labeling of venules in neurogenic inflamma­ Phannacol Chemother 31:138-151,1967 tion resembles that observed in the immediate response to inflam­ J ancso N, Jancso-Gabor A, Szolcsanyi J: The role of sensory nerve endings in neurogenic inflammation induced in human skin and in the eye and paw of the matory mediators such as histamine, bradykinin, and serotonin rat. BrJ P/larmacol Chemother 33:32-41, 1968 (Majno and Palade, 1961). As with these mediators, the plasma Jiinig W, Lisney S}W: Small diameter myelinated aJferents produce vasodilatation but leakage in neurogenic inflammation results from an increase in the not plasma extravasation in rat skin.J Physiol (Lond) 415:477-486,1989 vascular permeability, due to the formation of gaps between Karanth SS, Springall DR, Kuhn DM, Levene MM , PolakJM: An inununocytochem­ ical study of cutaneous innervation and the distribution of neuropeptides and endothelial cells (McDonald, 1994b). The endothelial gaps are protein gene product 9.5 in man and commonly employed laboratory animals. small (average diameter of 0.3 to 0.5 jLm) and are focally distributed AmJ Anat 191:369 -83,1991 around the endothelial cell borders. The gaps can be recognized by Kopaniak MM, Hay]B, Movat HZ: The elfect of hyperemia on vascular permeability. light microscopy after silver nitrate staining (McDonald, 1994a; and Microvasc Res 15:77-82,1978 Lembeck F: Sir Thomas Lewis's nocifensor sytem, histamine and substance P­ Fig 3E), by lectin staining (Thurston et aI, 1996), and by scanning containing primary aJferent nerves. Trends Neurosci 6:106-108,1983 and transmission electron microscopy (Hirata et aI, 1995; Baluk et Lembeck F, Holzer P: Substance P as neUrogenic mediator of antidromicvasodilation aI, 1997; and Fig 31'). The endothelial gaps are transient and fully and neurogenic plasma extravasation. Naunyn Schmiedebergs Arch Pharmacol 310: reversible, so that the plasma leakage ends in a few minutes after the 175-183,1979 Lewis T: The Blood Vessels of the Human Skin and their Responses. Shaw Sons, Ltd., stimulus (McDonald, 1994a; Baluk et aI, 1997). & London, 1927 Lou YP: Regulation of neuropeptide release from pulmonary capsaicin-sensitive alferents in relation to bronchoconstriction. Acta Physiol Scand 149:1-88,1993 Supported in part by a grantfrom the National Heart, Lung, and Blood Institllte Lundberg JM: Pharmacology of cotransmission in the autonomic nervous system: (HL-24136). integrative aspects on amines, neuropeptides. adenosine triphosphate, amino acids and nitric oJride. Pharmacol Rev 48:113-178,1996 Lundberg JM, Martling CoR, Hokfelt T: Airways, oral .cavity and salivary glands: REFERENCES classical transmitters and peptides in sensory and autonomic motor neUrons. In: Bjorklund A, Hokfeit T, Owman C (cds.). Handbook